Relation between critical speed and long-distance swimming performance: exploratory study.

This exploratory study aimed to evaluate the association of single- and multi-parameter models based on short-distance front crawl dash (i.e., 400 [T400] and 50 [T50] meters (m)) and critical speed (CS) with long-distance performance in talented swimmers. T400 and T50 of 26 females and 22 males (15.6 ± 1.6 years) were tested in a controlled setting and used to compute CS. Associations between long-distance swimming performance of 800 and 1500 m events and CS, T400, or T50 were examined, using single- (separate simple linear regressions) and multi- (multiple linear regressions with cluster-robust standard error) parameter models. Multi-parameter models' predictors (age, sex, D', rate of perceived exertion after 50 and 400 m) were selected using backward-elimination procedure. CS and T400 showed the greatest explanatory capacity in single-parameter models in both 800 (CS: adjusted R² [adjR²] = 0.802; T400: adjR² = 0.796; T50: adjR² = 0.383) and 1500 m performance (CS: adjR² = 0.807; T400: adjR² = 0.805; T50: adjR² = 0.454). Multi-parameter models showed similar model fit for both 800 (CS: adjR² = 0.807; T400: adjR² = 0.801; T50: adjR² = 0.802) and 1500 m (CS: adjR² = 0.808; T400: adjR² = 0.804; T50: adjR² = 0.805) performance. Multi-parameter models offered little improvement over single-parameter models (800 m: ΔadjR² = 0.005; 1500 m: ΔadjR² = 0.001), suggesting that CS or T400 alone may provide a practical and non-invasive tool for estimating long-distance swimming performance in talented athletes.